The authors are investigating reverse engineering for reconstructing the shape of simple mechanical parts. Currently, preliminary B-rep models can be created by fitting surfaces to point clouds obtained by scanning an actual part using a 3D laser scanner. The resulting model, although valid, is often not suitable for purposes such as redesign. This is because expected regularities and constraints are not present in the model. This report describes a number of aspects of the geometry of mechanical parts which should be exploited to adjust a B-rep model to improve its usefulness. Aspects considered are geometric constraints between surface parameters, regularly repeated substructures, symmetry, and the presence of features such as slots and holes. The results of a survey of a range of mechanical parts are presented and discussed, showing which of these aspects occur with a frequency that justifies their use in beautification algorithms intended to turn preliminary reverse engineered B-rep models into models engineers expect.

Boundary representation models reconstructed from 3D range data suffer from various inaccuracies caused by noise in the data and the model building software. Such models can be improved in a beautification step, which finds geometric regularities approximately present in the model and imposes a consistent subset of them on the model. Methods to select regularities consistently such that they are likely to represent the original, ideal design intent are presented. Efficiency during selection is achieved by considering degrees of freedom to analyse the solvability of constraint systems representing the regularities (without actually solving them). Priorities are used to select regularities in case of inconsistencies. The selected set of constraints is solved numerically and an improved model is rebuild from the solution. Experiments show that the presented methods can beautify models by selecting consistent regularities and enforcing major intended regularities.

Abstract. Perfect, partial, and approximate symmetries are pervasive in 3D surface meshes of real-world objects. However, current digital geometry processing algorithms generally ignore them, instead focusing on local shape features and differential surface properties. This paper investigates how detection of large-scale symmetries can be used to guide processing of 3D meshes. It investigates a framework for mesh processing that includes steps for symmetrization (applying a warp to make a surface more symmetric) and symmetric remeshing (approximating a surface with a mesh having symmetric topology). These steps can be used to enhance the symmetries of a mesh, to decompose a mesh into its symmetric parts and asymmetric residuals, and to establish correspondences between symmetric mesh features. Applications are demonstrated for modeling, beautification, and simplification of nearly symmetric surfaces. Key words: symmetry analysis, mesh processing 1

Current reverse engineering systems are able to generate simple valid boundary representation (B-rep) models from 3D range data. Such models suffer from various inaccuracies caused by noise in the input data and algorithms. The quality of reverse engineered geometric models can potentially be improved by finding candidate shape regularities in such an initial model, and imposing a suitable subset of them on the model by using constraints, in a postprocessing step called beautification. Finding such candidate regularities is a necessary first step, and is discussed in this paper. Algorithms for analysis are presented which use feature objects to describe properties of faces, edges and vertices, and small groups of these elements in a B-rep model with only planar, spherical, cylindrical, conical and toroidal faces. The methods seek similarities between feature objects, e.g. axes which are parallel, for each property type. For each group of similar feature objects they also try to find a special feature object which might represent the group, e.g. an integer value which approximates the radius of similar cylinders. The feature objects used represent shape parameters, directions, axes and positions present in the model. Experiments show that the regularities found by these algorithms include the desired regularities. Although other spurious regularities which must be discarded in subsequent beautification steps are also produced, their number can be reduced by appropriate choice of tolerance values.

Approximate geometric models, e.g. as created by reverse engineering, describe the approximate shape of an object, but do not record the underlying design intent. Automatically inferring geometric aspects of the design intent, represented by feature trees and geometric constraints, enhances the utility of such models for downstream tasks. One approach to design intent detection in such models is to decompose them into regularity features. Geometric regularities such as symmetries may then be sought in each regularity feature, and subsequently be combined into a global, consistent description of the model’s geometric design intent. This paper describes a systematic approach for finding such regularity features based on recovering broken symmetries in the model. The output is a tree of regularity features for subsequent use in regularity detection and selection. Experimental results are given to demonstrate the operation and efficiency of the algorithm.

Boundary representation models reconstructed from 3D range data suffer from various inaccuracies caused by noise in the data and by numerical errors in the model building software. The quality of such models can be improved in a beautification step, where geometric regularities need to be detected and imposed on the model, and defects requiring topological change need to be corrected. This paper considers changes to the topology such as the removal of short edges, small faces and sliver faces, filling of holes in the surface of the model (arising due to missing data), adjusting pinched faces, etc. A practical algorithm for detecting and correcting such problems is presented. Analysis of the algorithm and experimental results show that the algorithm is able to quickly provide the desired changes. Most of the time required for topological beautification is spent on adjusting the geometry to agree with the new topology.

Detecting approximate symmetries of parts of a model is important when attempting to determine the geometrical design intent of approximate boundary-representation (B-rep) solid models produced e.g. by reverse engineering systems. For example, such detected symmetries may be enforced exactly on the model to improve its shape, to simplify its analysis, or to constrain it during editing. We give an algorithm to detect local approximate symmetries in a discrete point set derived from a B-rep model: the output comprises the model’s potential local symmetries at various automatically detected tolerance levels. Non-trivial symmetries of subsets of the point set are found as unambiguous permutation cycles, i.e. vertices of an approximately regular polygon or an anti-prism, which are sufficiently separate from other points in the point set. The symmetries are detected using a rigorous, tolerance-controlled, incremental approach, which expands symmetry seed sets by one point at a time. Our symmetry cycle detection approach only depends on inter-point distances. The algorithm takes time O(n 4) where n is the number of input points. Results produced by our algorithm are demonstrated using a variety of examples.

Reverse engineering is the process of obtaining a geometric CAD model from 3D points acquired by scanning an existing physical model. It is widely used in numerous applications, such as manufacturing, industrial design and jewelry design and reproduction. We argue that for creating editable CAD models meant for manufacturing it is more appropriate to use feature-based constraint-based representations, since they capture design intent. We provide a framework for reverse engineering of small objects and in particular jewelry that combines cross section identification, feature and constraint information exploitation to attain robust, accurate and editable CAD models. First, we extract certain candidate features for describing our point cloud. These features are then reconstructed to describe the solid object. Constraints are automatically detected and maintained. Constraints capture design intent and provide robustness guaranties. Voxel inspired techniques are also employed to describe repeated patterns common to various types of traditional jewelry.

Motivated by the need to detect design intent in approximate boundary representation models, we give an algorithm to detect incomplete symmetries of discrete points, giving the models ’ potential local symmetries at various automatically detected tolerances. Here, incomplete symmetry is defined as a set of incomplete cycles which are constructed by, e.g., a set of consecutive vertices of an approximately regular polygon, induced by a single isometry. All seven 3D elementary isometries are considered for symmetry detection. Incomplete cycles are first found using a tolerance-controlled point expansion approach. Subsequently, these cycles are clustered for incomplete symmetry detection. The resulting clusters have welldefined, unambiguous approximate symmetries suitable for design intent detection, as demonstrated experimentally.

We aim to reconstruct three-dimensional polyhedra from axonometric line drawings. Existence of mirror symmetry in polyhedra can assist the reconstruction process. We present a new approach for determining planes of mirror symmetry of such polyhedral objects based on prior detection of their planar faces and any axes of symmetry of these faces. The axes are obtained from skewed facial symmetries, for which we also give a new method of determination.